Cristiano Padovani IGD-TP Exchange Forum (WG2) – 26 October 2016

Mechanical-Corrosion Effects on the Durability of High Level Waste and Spent Fuel Disposal Containers

2

Acknowledgments

•Fraser King (Integrity Consulting)

•David Sanderson and Paul Gardner (MMI Engineering)

•Sarah Watson (Quintessa)

Outline

• Introduction

• Quantitative analysis for carbon steel container in a bentonite-backfilled GDF subject to hypothetical loading regime

• Qualitative analysis of coated/cladded designs and general illustration of FADs to demonstrate container durability

• Summary

External loads in a GDF

External stresses arising due to: • Hydrostatic loads

• Lithostatic loads (for ‘plastic’ rocks)

• Buffer swelling (for clay-based buffers)

Evolution of mechanical stresses in a clay host rock

Courtesy of Nagra

Environmental conditions in a GDF

Changes in environmental conditions: • Cooling due to power decay

• Reduction in redox conditions due to oxygen consumption

Illustration of the evolution of T and Eh in a GDF

Courtesy of NWMO

Recent in-situ tests indicate shorter oxic period

Interaction between mechanical- and corrosion-related mechanisms

• Both corrosion and mechanical factors contribute to container failure

• Key failure modes for GDF – Plastic collapse due to loss of wall

thickness or increase in stress – Brittle fracture as a result of crack

growth or increase in stress – Plastic collapse and/or brittle

fracture due to degradation of material properties

– Through-wall penetration without loss of overall structural integrity

Mechanical factors

General corrosion

Hydrogen-related degradation

Failure modes Material and corrosion factors

Increased hydrostatic load due to glaciation

Buffer swelling pressure

Hydrostatic pressure

Internal pressurisation due to entrained water

Impact damage due to handling accident

Localised corrosion

Lithostatic loadfrom host rock

Impact damage from rockfall

Impact damage from seismic activity

Residual stress from fabrication

Shear load from seismic activity

Fatigue loading during transportation

Increased probability of plastic collapse as a result of loss of wall

thickness and/or increase in applied and residual

stresses

Increased probability of brittle fracture as a result of crack growth and/or increase in applied and

residual stresses

Increased probability of plastic collapse and/or

brittle fracture as a result of degradation of material

properties

Through-wall penetration without loss of overall

structural integrity

Brittle fracture due to cyclic loading

Stress corrosion cracking

γ-radiolysis effects

Neutron irradiation

Atmospheric corrosion

Corrosion fatigue

Galvanic corrosion

Microbially influenced corrosion

Thermal ageing

Creep

Fatigue crack growth

Failure Assessment Diagram (FAD)

• One method for conducting an engineering critical assessment for structures containing flaws, developed by UK nuclear industry (BS7910, CEGB R6)

• Demonstrates the proximity of a component to plastic collapse or brittle fracture

• Time dependence illustrated by a ‘trajectory’ of assessment points

Illustrative FAD curve

9

Quantitative analysis – carbon steel container

Conceptual design • Single-wall, min. wall thickness 75 mm (weld)

Hypothetical environmenta/loading scenario • Disposal in a high strength host rock in

compacted bentonite • Hydrostatic and buffer swelling loads • σY-magnitude residual stresses in closure weld • (Glaciation loads at 50,000 years)

Assumed flaws • Semi-elliptical internal or external weld flaws • Semi-elliptical internal flaw in base of

container at location of high stress • Extended full-circumferential internal weld flaw • Elliptical embedded flaws in weld

Conceptual disposal container option for AGR fuel (C-steel)

Assumed scenario

Specific assumptions Dry storage before disposal

- entrained water leads to minor internal corrosion but some H2 pressurisation

Disposal in GDF

• Mechanical evolution - hydrostatic pressure (7 MPa) - buffer swelling (6.5 MPa) - glacial loads at 55,000 years (18 MPa) - decrease in Kmat with time due to HIC

• Corrosion Behaviour

- corrosion rate = 1 µm year-1

P of H2 assumed in this analysis (equal to bentonite swelling P)

Illustration of time-dependent degradation of material properties (kmat)

• With bentonite buffer, anaerobic corrosion will lead to development of H2 (max P ~ Pswel + Phydro)

• Absorbed H leads to decrease in fracture toughness

• Flaw becomes “less safe” as the material properties degrade

Internal flaw (10 mm)

Example: circumferential internal weld flaw

• Assumed flaw depths of 10, 20, 30 mm

In the absence of glacial loads still substantial integrity after 10,000s of years (at 35,000 years wall loss due to general corrosion is 35 mm- about 50%)

Example: Embedded weld flaw (central)

• Defect sizes –a=5mm; c=10mm –a=10mm; c=20mm

For external and, to an extent, embedded defects corrosion ‘eats them away’ over time

Comparison of effects of flaw type and location (a = 10 mm)

glacial loads (55,000 years)

• without glacial loads, all defects up to 10 mm in size within envelope

• when glacial loads present, some internal flaw are typically outside envelope

no glacial loads (35,000 years)

16

Qualitative analysis – coated/cladded containers

Comparison of material properties • Same environmental and

loading scenario as before

• FAD used to qualitatively investigate same carbon steel container design with copper coating and titanium cladding

• Coating bonded to the substrate, so the strain is the same and the stress can be estimated as:

σcoat = (Ecoat/ECS)σCS

Scope of analysis

Copper Ti (grade 2 and 7)

Carbon steel

Effect of container design: Cu coating

Compared with the carbon steel substrate, based on assumed properties: • a cold-sprayed coating would have higher propensity to crack • an annealed coating (or annealing of a cold sprayed coating) would have a

higher resistance to crack

E (GPa) KIC (Mpa m0.5) Cu cold sprayed

120 130

Cu annealed 120 500 C-steel (no HIC)

200 220

0

0.2

0.4

0.6

0.8

1

1.2

0 0.5 1 1.5 2 2.5

K r

Lr

C-steelAnnealed CuCold spray Cu (as deposited)Cold spray Cu (as deposited)C-steel (AR)OFP Cu

Effect of container design: Ti cladding

Compared with the carbon steel substrate, based on assumed properties:

• a Ti-clad coating would have higher propensity to crack than the steel substrate • propensity to crack moves towards unsafe region with increasing [HABS] as KIH

decreases

E (GPa) KIC (Mpa m0.5) Titanium (no HIC)

105 50-80

C-steel (no HIC)

200 220

0

0.2

0.4

0.6

0.8

1

1.2

0 0.5 1 1.5

K r

Lr

C-steel

Ti Grade 2

Ti Grade 2 (1000 ug/g Habs)

Ti Grade 2 (700 ug/g Habs)

Ti Grade 2 (AR Habs)

C-steel (AR)

Summary • Assessing the coupling between corrosion and mechanical processes

on the durability of HLW/SF containers, can be important in the case of: – Crack growth supported by increased load due to wall thinning – Degradation of mechanical properties (due to H2 absorption,

radiation embrittlement, …)

• Failure Assessment Diagrams can be used to illustrate the evolution of container integrity and its proximity to failure

• Beyond a qualitative analysis, the approach can be used qualitatively to

evaluate the implications on container design on the likely durability

Back-up slides

Effect of host rock type High-strength rock (HSR)

• Reference trajectory of assessment points described above

Low-strength rock (LSR) • Similar trajectory but

container would likely fail earlier because of additional lithostatic load

Evaporite • Absence of water leads to

less wall loss and H absorption

• Lower tendency to brittle fracture and slower approach to plastic collapse

Assumes same corrosion mechanisms and rates

Effect of buffer type

• Lower rate of general corrosion leads to reduced wall loss and lower [HABS]

• Overall, lower probability of brittle fracture and slower progress towards plastic collapse

Assumed flaws in container 10-mm-deep hemispherical flaws located circumferentially at either the inner or outer surface of the final closure weld or radially in base of container in region of high tensile stress

Embedded flaw in closure weld (quarter)

Defect sizes • a=5mm/c=10mm • a=10mm/c=20mm

Embedded flaw in closure weld (inner)

Defect sizes • a=5mm/c=10mm • a=10mm/c=20mm

Sensitivity to internal flaw size

Flaw sizes of 10-30 mm after 35,000 yrs

Internal loads in a disposal container

Potentially, internal stresses due to: • Water/hydrogen gas (corrosion and radiolysis)

• Helium gas (α-decay)

Evolution of internal stresses in a disposal container due to entrained water (spent AGR fuel, assumed to containing 1.4kg of residual water)

Time (y)

Pres

sure

(MPa

)

2035 2025

0.25

0

High anaerobic corrosion rate High aerobic corrosion rate Reference corrosion rates

Diapositive numéro 1�Cristiano Padovani�������IGD-TP Exchange Forum (WG2) – 26 October 2016AcknowledgmentsOutlineExternal loads in a GDFEnvironmental conditions in a GDFInteraction between mechanical- and corrosion-related mechanismsFailure Assessment Diagram (FAD)Diapositive numéro 9Assumed scenarioSpecific assumptionsIllustration of time-dependent degradation of material properties (kmat)Example: circumferential internal weld flawExample: Embedded weld flaw (central) Comparison of effects of flaw type and location (a = 10 mm)Diapositive numéro 16Comparison of material propertiesEffect of container design: Cu coatingEffect of container design: Ti claddingSummaryBack-up slidesEffect of host rock type�Effect of buffer type�Assumed flaws in containerEmbedded flaw in closure weld (quarter)Embedded flaw in closure weld (inner) �Sensitivity to internal flaw size�Internal loads in a disposal container